Saturday 5 March 2016

MUSINGS: Cable Claims, Testimony ("Buyer's Guide") & Sponsored Content...

Cute CD BTW for the Simpsons lovers out there...
I must admit that I rarely pay attention to the folks at Hi-Fi+. Other than occasionally skim over an issue at the magazine stand, I find it generally lacking in details and looks too much like one big glossy advertisement.

Shortly after writing the previous article on Chromecast Audio including jitter discussions, I noticed on Dr. AIX's recent post that he linked to a 2015 "Hi-Fi+ Guide To Cables" PDF and decided to wander over for a peek (I know... Bad move, right?). I'm not going to enumerate all the disturbing comments and beliefs advocated but I would like to point the reader at the modestly interesting interviews with the various founders and representatives of the cable companies. Particularly, there was one question asked of each manufacturer - whether they would comment on the best type of digital connection to use: USB, I2S, Ethernet, or coaxial S/PDIF (why not include TosLink S/PDIF?).

A number of manufacturers appear to have no comment (eg. Black Rhodium, Audience), non-committal (eg. Kimber), or answered the question tangentially (eg. Synergistic, Chord). A few had pretty good responses and appeared aware of the technical benefits of I2S (eg. AudioQuest, Vertere). But I found it interesting how Nordost recommends S/PDIF "for best results" based on impedance reasons. And MIT likewise recommended S/PDIF because they somehow have "so much control" over signal integrity! Really? These people do realize that S/PDIF is more than likely the most jittery of all the options provided and the coaxial variant also doesn't provide galvanic isolation, right? I thought jitter was a "bad" thing. Perfect impedance match itself isn't going to affect jitter in a meaningful way no matter how much "control" they think they have in terms of construction of the cable.

Even more curious, consider the response from Crystal Cables / Siltech which I found to be the most disappointing. Supposedly USB and ethernet have wire configurations causing "major jitter and picks up noise". Really? And ethernet "fails as their standard connection is poorly implemented mechanically in regard to impedance and jitter". Really!? So sad... And I was starting to like ethernet after demonstrating signal integrity over >7,300km with essentially no measurable jitter :-).

Would any of these companies please demonstrate an example of how their cable improves quality and presumably reduce the effect of jitter on actual DAC analogue output compared to a decent run of generic cable? Maybe MIT and Nordost can give it a go with their coaxial S/PDIF cables which they have mastery in? Crystal Cables and Siltech make USB cables; how about showing the benefits of their USB compared to the "major jitter" from generic cables?

I wonder if the cable manufacturers have been scared off ethernet due to all the bad press around the "snake-oil" nature of their claims especially last year. Good. I hope they stay far, far away and over time perhaps we'll have the good fortune to see other cable types fall out of favour from silver-tongued hype.

Although much of the comments up to now have to do with the recent Hi-Fi+ "buyer's guide", let's think about this in the broad context and consider the other magazines like Stereophile, The Absolute Sound, What Hi-Fi?, etc... When was the last time writers actually did any controlled evaluation, or questioned the value of expensive cabling even though it's not hard to find consumers and websites like this blog questioning the relevance of the whole "audiophile cable" venture?

I was listening to a radio program a couple weeks back about the evolution of media in general from traditional print-based models to how things have shifted these days with Web-based delivery. One general observation made has been that "investigative journalism" has suffered over the years. So often these days (and not just in audiophilia), articles have become brief announcements or soundbites containing nothing more than product news aimed at garnering "traffic", and often the faster the story is posted, the more clicks it receives as the first to announce. Even with more substantial "review" articles, notice just how much text is devoted these days to comments from manufacturers that look like nothing more than cut-and-paste jobs with much of the content little more than what could be garnered by looking at a manufacturer's website! Ralphpnj on the Squeezebox forum recently pointed to an article describing the rise of "sponsored content" in news media. Indeed, let's just be candid and acknowledge that so much of what we read in the audio world is simply "sponsored content" (obviously that was the nature of the free "guide to cables" above).

Consider for a moment when was the last time you saw a magazine article go in depth into exploring a commonly used word like "jitter". If I just started exploring high-end audio in the last 6 months, reading magazines, joining the typical audio forums, do you think a "newbie" would understand this concept or would he end up confused, perhaps even a little anxious with doubts as to the adequacy of jitter control in his own system? Googling "Stereophile jitter" (giving Stereophile the benefit of being perhaps the most rational of the bunch), he would run into articles from 2004-2009 rather than contemporary discussions. I would hope to think that this state of affair is not something that is acceptable for serious hobbyists nor should ignorance be an acceptable status quo for the press.

About a year ago, I wrote the post "Audiophiles 'Us vs. Them' (Objectivists vs. Subjectivists)". In it, I expressed that from the perspective of how I feel about all this, I'm actually not angry. And I find it hilarious that anyone would even think envy is a significant factor among the "more objective" audiophiles. If anything, I suppose there is disappointment mixed in with disgust. Disappointment with the lack of substance when reading purely subjective reviews and thinking that there's more we could know about a device from the accuracy perspective if only we could be treated with some objectivity to demonstrate just how much effort went into the engineering. And disgust when comments are made by some like these manufacturers above who speak with conviction but not truth. I fully accept that we all have idiosyncratic preferences. Sure, some might prefer vinyl over digital, or tubes vs. solid state, or jazz vs. classical vs. rock vs. pop vs. electronica... But concepts like accuracy and jitter do have objective correlates which must be respected; otherwise the endeavour for something like an electronic hardware review would be incomplete. Without grounding and some form of reality testing, one might as well enter the Twilight Zone or fantasy world where truth and opinion end up undifferentiated. With the audiophile media space lacking in consumer advocates and educators the unfortunate outcome of "sponsored content" is that magazines these days seem to accept on faith claims made by these sponsoring sources no matter how incredible!

I have heard apologists say that "this is only a hobby". And aren't hobbies meant to be "pleasurable"? Of course. But since when are half-truths and at times all-out lies acceptable and ethical conduct, not to mention that many of these instances seem to be part of typical business practice?

Let me end off with a practical challenge to audiophile magazines to perhaps help readers appreciate the effect of jitter. In the article on digital audio by Julian Dunn linked previously in my Chromecast Audio Part II post, consider the following summary from page 34:

Audibility considerations
It is one thing to be able to identify and measure sampling jitter. But how can we tell if there is too much?

A recent paper by Eric Benjamin and Benjamin Gannon describes practical research that found the lowest jitter level at which the jitter made a noticeable difference was about 10 ns rms. This was with a high level test sine tone at 17 kHz. With music, none of the subjects found jitter below 20 ns rms to be audible. 
This author has developed a model for jitter audibility based on worst case audio single tone signals including the effects of masking. This concluded:

“Masking theory suggests that the maximum amount of jitter that will not produce an audible effect is dependent on the jitter spectrum. At low frequencies this level is greater than 100 ns, with a sharp cut-off above 100 Hz to a lower limit of approximately 1 ns (peak) at 500 Hz, falling above this frequency at 6 dB per octave to approximately 10 ps (peak) at 24 kHz, for systems where the audio signal is 120 dB above the threshold of hearing.”

In the view of the more recent research, this may be considered to be overcautious. However, the consideration that sampling jitter below 100 Hz will probably be less audible by a factor of more than 40 dB when compared with jitter above 500 Hz is useful when determining the likely relative significance of low- and high-frequency sampling jitter.
(Bold added to highlight important details!)

Perhaps one of the audiophile magazines could take on the task of unpacking the above information and presenting it to the general public. Maybe do some research on what has been done since 2003, perhaps provide some illustrations... Maybe show some J-Test results to explain what the audible thresholds might look like in equipment reviews. I suspect a well done article like that would really become a "primer" of sorts for audiophiles; likely quoted, recommended, and searched for years to come. Imagine how many clicks and how much traffic an article of substance could garner over the years. The impact would be very significant in demystifying what should have never been that much of a mystery to begin with (remember we're looking at the impact of jitter in audio frequencies over wires, not gigahertz wireless telecommunications!). It might even help some of these cable manufacturers understand and test a thing or two before they speak. Of course that's assuming that education and reality testing are within the mandate of magazines these days and that there is editorial freedom to explore beyond the limits of sponsored content or desires of said sponsors.

-------------------------

The other day, I was with my 10 year old son at the local bookstore to see if there was anything he wanted to pick up. Among the usual comics and youth novels, he picked up a book on "cryptozoology"; essentially the "monsters" of this world (even though of course some could be just undiscovered creatures rather than scary). You know, Bigfoot, Wendigo, the chupacabra. As I was thumbing through the book with my son, I could not help but smile and realize that there was more "objective evidence" in the form of fuzzy pictures, footprint casts, shaky videos, strange animal tracks ostensibly for Bigfoot, Yeti, and the Loch Ness Monster that I have ever seen as evidence for the benefits of expensive cabling on sound quality. Testimony of course is just as prominent if not more in the audiophile world. Go figure... :-)

Have a great week ahead. Enjoy the music everyone!

15 comments:

  1. SPDIF via RCA can be isolated by the way but generally isn't. Inputs or outputs using a coupling transformer are galvanically separated but still have a poor HF damping. One (or more) ferrites can help here.

    Jitter (and cables) are BIG business where lots of money IS (easily) made.
    The majority of people really WANT to believe everything related to it.
    Their ears tell them they are right too.
    These people want to be educated but not by non-believers, they merely seek verification and there is LOTS of that around.

    So keep educating is all I can say.

    ReplyDelete
    Replies
    1. Hi Frans.

      Indeed, that the idea of being "preached to by the choir" is very much a problem here, isn't it?

      Delete
  2. Wonderful to read this. I'm a fellow audio engineer and absolutely disgusted by the direct lies the snakeoil companies are telling and the corruption of audiophile magazines. The fact is that if they don't meaure them in any way, it is not crdible review. If nothing else, to confirm that the specs are what manufacturers says they are.. Oh wait, these aholes, AQ, nordost etc do not even give out gauges on speaker cables! They are not rated, we have no idea what temperature it can handle, basic, simple data. If we can't use it on commercial installation due to safety concerns, they are NOT high end. The head of AQ said to me that they don't give out datasheet since that is trade secrets.. lol... I'm in a fight against the scammers.

    ReplyDelete
  3. "Perhaps one of the audiophile magazines could take on the task of unpacking the above information and presenting it to the general public"

    But ... that would involve writing an article that said the readers _didn't_ have to worry about using fancy ultra-low-jitter connections! I think we can all see the problem here. And Heaven forbid they mention that all this jitter stuff is irrelevant as a decent signal receiver will reclock the data anyway (which is why you can get zero jitter from a 7,300km stretch of network cable). Maybe they should think about switching to pro-audio instead, where jitter might actually matter as you are using long cable runs and can have mutiple pieces of equipment that all needs to run on the same clock.

    One nitpick though: impedance mismatches will affect jitter on a coaxial transmission-line interface as it causes reflections that bounce back-and-forth and smear the signal. There's an article by Steve Corrigan from TI that explains this: http://www.ti.com/lit/an/slyt179/slyt179.pdf

    ReplyDelete
    Replies
    1. Thanks Charles. Yes.

      It could be extremely dangerous to have the "inconvenient truth" be revealed that for the *vast majority* of cases, cables in an audiophile system may just make no difference to jitter performance. No magazine would *ever* want to make advertisers feel bad about the "performance" of their products or whittle away even more of the faithful few who still believe in all this nonsense after all!

      Bad for magazine/website revenue. Bad for sales. Bad for the used market. Bad to take away the sense of this boutique stuff being any kind of "investment" with enduring value.

      Thanks for the TI article. Yes, I have seen articles like this and have come across various discussions as well... However, despite the possibility and arguments, I have yet to see demonstrations as to the effect of impedance on the analogue output whether in the J-Test or other measurements as it applies to S/PDIF cables.

      Remember this test I did back in 2013:
      http://archimago.blogspot.ca/2013/04/guest-review-measurements-dr-franks.html

      Despite the horror of that "Frank's Best" "coaxial" cable (nothing of the sort of course!)using a cheap CM6631A USB to S/PDIF converter and inexpensive eBay/Chinese AUNE X1 DAC, I could not demonstrate any change with the Dunn J-Test! Like I said, I would love if these companies would just show us *one* reasonable example of how their cable bests a "normal" cable so we can understand the value of their claims...

      Delete
  4. Brilliant articulate article!
    Coaxial SPDIF cables all sound different because the cable impedance varies and the connector and receiver impedance varies. This means that some cables will match better with some equipment than others. But ultimately Toslink is better because it provides galvanic isolation. I think the cable industry relies on the concept of "a change is as good as a holiday", because they know that their cable will sound different because of slight variations in impedance. Ultimately if there is a difference in sound then people are using the wrong connection. Digital connections shouldn't sound different!

    ReplyDelete
  5. Question 1: Does a power cleaner make any audible difference? Apart from cables, Hifi dealers want to sell you such devices. I use a 220V amp with a transformer ($35 at the time) to bring the voltage down to 110 V and my amp sounds the same in Europe and Canada. Despite the different alignment of air molecules on either side of Greenwich. But if I used a real power cleaner, would it eliminate the effects of the Coriolis force?

    Question 2: Do you need a surge protector to protect your equipment? I do have a bunch of them that had been accidentally sent to me by Belkin. They let me keep them. If they didn't offer them, I would not have these freebies.

    Question 3: I have used 12 gauge copper cables for my speakers since I bought them in 1985. Shouldn't that be optimal?

    Question 4: Standard outdoor extension chords for my Christmas lights...will they do as decent speaker cables when cut appropriately?

    ReplyDelete
  6. I'm game...

    Answer 1: It shouldn't unless something is really wrong with the mains power supply (lots of transients etc) or when the equipment used is highly susceptible to nasties riding on the mains and there is lots of it that it becomes audible.

    1b: Will electrons travel relatively faster near the equator and is this audible ?

    Answer 2: When you live in an area where lightning often strikes or sudden peaks happen in the mains grid, then a surge protector MAY protect some of the equipment unless the lightning strikes really close....
    Can't hurt to use them... they do NOTHING when the mains has normal voltages... well not entirely true technically as the varistors inside do have some small capacitance (between 100pF and 2nF) so could lower RF signals riding between Live and Neutral wires.
    When the surge protectors are connected to 3 prong sockets they could potentially give somewhat more protection IF there are also varistors present between Neutral and safety ground internally...

    Answer 3: Good enough for me, maybe strip the wire ends again if the wires have oxidised a bit.

    Answer 4: probably, may depend on the final length, the speakers connected and the required SQ.

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. Wow that's a wonderfull blog having all details & helpful. China Power Cord

    ReplyDelete
  9. I hope they stay far, far away and over time perhaps we'll have the good fortune to see other cable types fall out of favour from silver-tongued hype.Consumerism Inc.

    ReplyDelete
  10. Archimago,

    Please do not add to the perception that SPDIF (and it's cousin AES) is inadequate for audio. It is perfectly fine. You point out that it is technically inferior without much qualification at the top of this article.

    I have the bad feeling that SPDIF and AES interfaces are being squeezed out ironically because they make less money for placebo dealers. It's far more profitable nowadays to sell endless USB and I2S upgrades because those systems are inherently more complex, require computerized negotiation, etc, etc.

    I feel the squeeze myself because my aging collection of DSP devices which use AES seem headed toward end of life without replacement. Audiophile winds are strongly blowing against SPDIF/AES, anyone who claims to be an audiophile now MUST be using I2S so it is said. That means, pretty much, the end of freewheeling plug and play, people will be locked into proprietary computer programs, and I strongly resent and will resist this as hard as I can.

    I am glad the audiophile cable makers are mostly on my side on this. I'm not as angry about unprovable upgrades than you and many objectophiles are. I'm fine with a slightly more expensive silver coated and teflon insulated cable, so long as it meets the basic specifications ok. I don't think everything needs to be made as cheaply as possible.

    ReplyDelete
    Replies
    1. Hi AI,
      You're right, SPDIF and AES are actually perfectly fine! There is simply more variability in the jitter and as Solderdude below notes, it's really in the implementation.

      Delete
  11. SPDIF works fine in most cases but the implementation of the receiver is important.
    The disadvantage of SPDIF is that it is a combined signal (clock + data) where the clock has to be recovered.
    This means that bandwidth limited induced jitter is lurking and can affect the retreived clock jitter. Again ... very important ... it's the actual DAC's receiver and how they handle this that is what determines the final jitter at the DAC-chip inside the DAC.

    The same is true for AES but because it is balanced in nature the cable induced jitter is much smaller at the receiver side. That is ... depending on HOW the receiver circuit is constructed. Here too ...how this affects the actual jitter of the DAC-chip depends on the implementation of jitter reduction in the DAC itself (not the DAC-chip itself).

    I2S is primarily designed for 'communication' between chips on a single DAC board istelf. It is designed for SHORT wiring distances and needs to meet certain specifications to work well. As this is connected directly to the DAC chip itself (in most cases) it is not wise to use long wires and connectors in between.
    The fact that it still works well even when longer connection wires are used is a testimony to the protocol and its implementation rather than it being a 'better' interface. It does have a separate clock so does not have to be recovered.

    Usually there is no galvanic separation in I2S, which is possible with SPDIF and AES, so ground loops and common mode currents/nasties have more chance of doing harm.

    USB is a packet 'wrapped' and is not 'real time'. Clock regeneration is done differently here. Again... this highly depends on how the manufacturer implemented the signal handling and whether or not galvanic separation is used or not.

    How this all translates to audiophiles and what they perceive does not seem to be related to measured performance and actual mode of data transfer but more to the usual 'causes'.
    I realise that many will fully disagree here because many people can clearly 'hear' differences.
    So be it....

    I agree that cables don't need to be manufactured as cheaply as possible but don't think they should cost a lot more than say $20.- to $50.- to perform optimally either.

    ReplyDelete